Peetz0r/gist:91f484a97cc1128d7250

## gistfile1.py
#!/usr/bin/env python2

from __future__ import print_function
from pdfminer.converter import TextConverter
from pdfminer.layout import LAParams
from pdfminer.pdfpage import PDFPage
from pdfminer.pdfinterp import PDFResourceManager, PDFPageInterpreter
from calendar import monthrange
import StringIO, re, glob, os, pygal, locale, datetime, sys

locale.setlocale(locale.LC_ALL, '')

all_months_data = []
all_months_rendered = []

filenames = glob.glob("FACT*_*.pdf")
filenames.sort()

out  = '<!doctype html>'
out += '<html>'
out += '<head>'
out += '<meta charset="utf-8">'
out += '<title>Mobile data usage by day and by month</title>'
out += '<style>'
out += '* {'
out += 'background-color: black;'
out += 'padding: 0;'
out += 'margin: 0;'
out += 'text-align: center;'
out += 'line-height: 200%;'
out += 'color: white;'
out += 'font-family: sans-serif;'
out += '}'
out += 'svg {'
out += 'margin: 50px;'
out += '}'
out += '</style>'
out += '</head>'
out += '<body>'

for file_nr, filename in enumerate(filenames):
	print('[%s>%s]' % ('='*file_nr*2, ' '*(len(filenames)*2-file_nr*2)), end='\r')
	sys.stdout.flush()

	rsrcmgr = PDFResourceManager()
	retstr = StringIO.StringIO()
	codec = 'utf-8'
	laparams = LAParams()
	device = TextConverter(rsrcmgr, retstr, codec=codec)
	interpreter = PDFPageInterpreter(rsrcmgr, device)

	fp = file(filename, 'rb')
	for page in PDFPage.get_pages(fp):
		interpreter.process_page(page)
	fp.close()
	device.close()

	m = re.findall('Internet(\d{2}-\d{2}-\d{4})\d{2}:\d{2}:\d{2}\s*(\d+) kb', retstr.getvalue())
	retstr.close()

	total = 0.0
	now = datetime.datetime.strptime(m[0][0], '%d-%m-%Y')
	data_list = [0]*monthrange(now.year, now.month)[1]

	for day in m:
		day_nr = datetime.datetime.strptime(day[0], '%d-%m-%Y').day-1
		data_list[day_nr] += int(day[1])/1024
		total += int(day[1])

	all_months_data.append([now.strftime('%b%y'), round(total/1024/1024, 2)])

	bar_chart = pygal.Bar(width=1000, height=500, explicit_size=True, label_font_size=12)
	bar_chart.title = now.strftime('Mobile data usage by day in %B %Y (total: '+'%.2f GB' % (total/1024/1024)+')')
	bar_chart.x_labels = map(str, range(1,len(data_list)+1))
	bar_chart.add('MB', data_list)
	all_months_rendered.append(bar_chart.render())

all_months_rendered.reverse()

bar_chart = pygal.Bar(width=1000, height=500, explicit_size=True, x_label_rotation=30, label_font_size=14)
bar_chart.title = 'Mobile data usage by month'
bar_chart.add('GB', list(map(list.pop, all_months_data)))
bar_chart.x_labels = list(map(list.pop, all_months_data))
out += bar_chart.render()
out += ''.join(all_months_rendered)
out += '<hr>Generated at '
out += datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')
out += '</body>'
out += '</html>'

with open('mobile-data-usage.html', 'w') as fp:
	fp.write(out)

print('[%s]' % ('='*(len(filenames)*2+1)))
	#!/usr/bin/env python2

	from __future__ import print_function
	from pdfminer.converter import TextConverter
	from pdfminer.layout import LAParams
	from pdfminer.pdfpage import PDFPage
	from pdfminer.pdfinterp import PDFResourceManager, PDFPageInterpreter
	from calendar import monthrange
	import StringIO, re, glob, os, pygal, locale, datetime, sys

	locale.setlocale(locale.LC_ALL, '')

	all_months_data = []
	all_months_rendered = []

	filenames = glob.glob("FACT_.pdf")
	filenames.sort()

	out = '<!doctype html>'
	out += '<html>'
	out += '<head>'
	out += '<meta charset="utf-8">'
	out += '<title>Mobile data usage by day and by month</title>'
	out += '<style>'
	out += '* {'
	out += 'background-color: black;'
	out += 'padding: 0;'
	out += 'margin: 0;'
	out += 'text-align: center;'
	out += 'line-height: 200%;'
	out += 'color: white;'
	out += 'font-family: sans-serif;'
	out += '}'
	out += 'svg {'
	out += 'margin: 50px;'
	out += '}'
	out += '</style>'
	out += '</head>'
	out += '<body>'

	for file_nr, filename in enumerate(filenames):
	print('[%s>%s]' % ('='file_nr2, ' '(len(filenames)2-file_nr*2)), end='\r')
	sys.stdout.flush()

	rsrcmgr = PDFResourceManager()
	retstr = StringIO.StringIO()
	codec = 'utf-8'
	laparams = LAParams()
	device = TextConverter(rsrcmgr, retstr, codec=codec)
	interpreter = PDFPageInterpreter(rsrcmgr, device)

	fp = file(filename, 'rb')
	for page in PDFPage.get_pages(fp):
	interpreter.process_page(page)
	fp.close()
	device.close()

	m = re.findall('Internet(\d{2}-\d{2}-\d{4})\d{2}:\d{2}:\d{2}\s*(\d+) kb', retstr.getvalue())
	retstr.close()

	total = 0.0
	now = datetime.datetime.strptime(m[0][0], '%d-%m-%Y')
	data_list = [0]*monthrange(now.year, now.month)[1]

	for day in m:
	day_nr = datetime.datetime.strptime(day[0], '%d-%m-%Y').day-1
	data_list[day_nr] += int(day[1])/1024
	total += int(day[1])

	all_months_data.append([now.strftime('%b%y'), round(total/1024/1024, 2)])

	bar_chart = pygal.Bar(width=1000, height=500, explicit_size=True, label_font_size=12)
	bar_chart.title = now.strftime('Mobile data usage by day in %B %Y (total: '+'%.2f GB' % (total/1024/1024)+')')
	bar_chart.x_labels = map(str, range(1,len(data_list)+1))
	bar_chart.add('MB', data_list)
	all_months_rendered.append(bar_chart.render())

	all_months_rendered.reverse()

	bar_chart = pygal.Bar(width=1000, height=500, explicit_size=True, x_label_rotation=30, label_font_size=14)
	bar_chart.title = 'Mobile data usage by month'
	bar_chart.add('GB', list(map(list.pop, all_months_data)))
	bar_chart.x_labels = list(map(list.pop, all_months_data))
	out += bar_chart.render()
	out += ''.join(all_months_rendered)
	out += '<hr>Generated at '
	out += datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')
	out += '</body>'
	out += '</html>'

	with open('mobile-data-usage.html', 'w') as fp:
	fp.write(out)

	print('[%s]' % ('='(len(filenames)2+1)))